Gigabyte X570 Aorus Master won't power on - (temporary) solution

QUESTION? Did any of you have this problem before Bios F11? I don't recall having this problem with any previous bios, but it has happened twice in the last 2 months. I'm willing to bet it's a bios issue.

I never upgraded to F11 I had it in my TODO list. I am not sure what revision I was before the failure. After removing the battery its on version F10c.
 
Have the same Problem as you guys. So far it happened three times. After writing back and forth with Customer Service and them being not very helpful one of them send me the F12e Bios to try out. Was just wondering if any of you got access to it as well and tried it already before update. I did attach the File they sent but I'm obviously not gonna take responsibility for it. Probably wouldn't recommend flashing to a Bios Version from a random Guy on the Internet anyway. But in case someone want's to try it here it is.
Hi Luca9751, it's been a while and wondered if you have been error free since using the F12e bios.
I may give this a go next, my Son's PC has been running fine since the last battery pull on May 4th, no real changes to bios except putting xmp back on and plugging his PC directly into a ring main socket ie not into a power strip.
See how it continues.
 
Experienced cold boot issue from the beginning (ONLY on the first boot AFTER psu was switched off/unplugged) , had sudden reboots every now and then (sometimes 3-4 weeks without trouble),
but only when running office/chrome, never under idle/low load (streaming etc) or gaming/full load, so i put it down to bios/settings or maybe too much stuff i "tested" on my win install (wont reinstall until the new build gets released).
Until this morning when it wouldn't do anything, but noticed the boards rgb stuff wouldn't flicker when i switched the psu off/on.
It wasnt the case switch (Tt S300 has a crappy one), and psu would turn on just fine (evga 650w G+).

Didnt want to deal with cmos/battery, as i have 2 rads/7 fans, and vertical setup for pci stuff, so PITA to get to it, but after a quick google search to "double check" before removing things,
i found this thread, and wanted to comment on a few things as well.

YES, removing the battery and bleeding any residual charge on the rig (+15 min after holding case button for 30s) fixed it,
but that is only a temporary a fix.
Like taking pain killer for a broken bone (pain gone, but the broken bone is not).



I can basically eliminate everything else as cause, but the board.
Same ram/psu/drives/cooler/cpu were working for +6 month with a different board (MSI gaming edge),
and except the cpu, all other parts had run with an intel for about 6 month prior, so it can only be the board.

Problem seems to be more common with Master/Ultra boards (not the one above/below), so might be related to components on the board/layout?!

Im running F12e since it came out, and the first thing was a cold boot, so definitely NOT the fix.
And its NOT a requirement to unplug the rig after bios update, clearing cmos with button/jumper is usually enough.
Best proof is the ryzen lineup from Gb with at least the x570 retaining old bios settings, and not clearing the cmos after update,
neither thru shortening the pins and/or removing cmos battery etc.
only thru flashing the bios update, then (powered down) using recovery flashback (and same "new" bios file), and reflashing same update again once i was back in bios (including backup chip),
could i get it to work properly (last 3 bios updates).


Its not ram related, as it did it with 3 different kits, two gskill kits (on QVL) with samsung b die would barely run spd (3600-18/22/2242/[email protected]) and wouldn't even post unless i set voltage to 1.35 prior to switching from corsair,
which isnt even qvl, and even that it has "just" decent micron chips it does 3600-16/19/19/36/[email protected] (spd 18/22/22/42/[email protected]).
Ram is either stable or its not, so except for dying chips, not the cause, nor does anyone need to mess with trying different slots/amount of sticks etc (unless wrongly set up from the start)


Im using a UPS with AVR and true sine wave output, so i can exclude power coming in.

If its really "draining" the battery, this should happen more gradually/get worse over time or happen suddenly (full->discharged),
its not like i can have a drained battery causing sudden reboots multiple times per week, and then not have any for weeks, with same "drained" battery).



I should have known better after what Gb started a couple of years ago. they would release hw with R1.0, this would get reviewed/tested/benched and maybe even perform good on ocing (e.g. samsung chip for vram),
only to switch to lower priced components for R1.1/2.0 (e.g. hynix/micron ram) without mentioning it ANYWHERE (not even in the fine print), just the typical "we reserve the right to make future changes...").
Especially with other brands not doing so, was the main reason not to use Gb boards (first) or gpus anymore, nor recommend them.

To anyone know gong "ah, can happen/thousand of the same board run fine..": i build probably 50 rigs in the past 18y (myself/friends/associates),
and not one board had these issue/needed rma/replacement, and ALL worked for + 5-7y in one or even two different builds, including boards that were just 100$,
all performed better than my Ultra.
Outside the fact that i didnt spend +300$ on a board to have to remove half my hw and the cmos bat every now and then, just to get it to boot.
 
I've used every bios since the one the mboard shipped with back last july 09, when I installed it, F3. On that date I also installed a 3600X. Never had a problem with all bioses--used every one--through F11/F12a/ to my current F12b.... Never had a single problem up to the installation of my 3900X. Didn't have the first problem even then until a couple of days (3 days after, imo--cold-boot on the 4th day) after I installed my 3900X. I had two problems, initially, I now believe:

1) Cold-boot problem: Apparently caused by a bad/faulty pwr strip that didn't show a fault until I doubled the cold-boot PSU load on the strip by installing the 3900X in mid-March. That's when both my problems started. Cold-boot, battery drain problem was completely cured, immediately, by the installation of a newer, non-faulting, better power strip, https://www.amazon.com/gp/product/B000J2EN4S/ref=ppx_yo_dt_b_asin_title_o02_s00?ie=UTF8&psc=1 on May 2 (It's just an ordinary pwer strip, imo.) My cold-boot/battery drain problem has not recurred since! I am happy to report.

2) Temp motherboard brick: Shortly thereafter, however, after a simple software shut-down via Win10, IIRC, the mboard temporarily bricked (see my earlier posts for detailed descriptions of events 1 and 2)--no lights, no power, no nothing. First time this happened since the mboard install in July '19. After pulling the battery for 45 minutes, I was glad to see, system booted normally immediately! The problem has not recurred, thankfully, in 10 consecutive cold-boots and Win10 software shutdowns! (BTW, I disable sleep/hibernation--never use 'em--just a habit I got into years ago.)

Analysis: When I swapped my 3600X for a 3900X, I should have pulled the battery, too, just out of proper technique... I'm lazy sometimes and thought that resetting CMOS via the backpanel would be sufficient. Likely it wasn't, obviously. I think that the original NVRAM corruption (as AZRAK was kind enough to mention in this thread!) was partially caused by the increased power demand between the 3600X and 3900X during cold boot--the 3900X doubling that demand during cold boot, essentially--coupled with my bad/faulty/failing/insufficient pwr strip--which was sufficient to corrupt the NVRAM a bit more(!)--which then created the temp mobo brick condition--indicating further NVRAM corruption. --So, pulling the battery this time, after having corrected the power strip issue, was sufficient to solve the NVRAM boot problem, permanently. Here's hoping...!

Conjecture: During event #1, I pulled the battery, and then I replaced the battery, when the simple pull of the original battery didn't work as the cold-boot failure resumed. Likely, the original battery may have been fine. It wasn't until I began thinking about the pwr strip and then replaced it that the cold-boot failures ceased, as well as the battery drain. Event number #2 was likely caused by residual NVRAM data corruption remaining after the cold-boot failure problem was solved. And so, a final battery pull was required in order to clear this last bit of bad NVRAM data that had accrued.

If this is all correct then I should have no more problems with the mboard! We'll see! I certainly hope so. I have this thread marked and so I will certainly return if something should fail along these lines in the future...;) Although I've not had these exact problems before--I've also never owned a 12-core/24t CPU before, either...!
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
Glad you got it running again, but there is no battery "drain" issue.
Im still using the same battery it came with, now for almost 8 month, so it's obviously not being discharged,
yet i still have the same probs as others here (pc wont POST after i switched psu off/unplugged power, cold boot here and there even without psu off, sudden reboots..).

Its all related to the dual bios stuff. I found post this is going on since Gb switched from single to dual around 2000,
and is not limited to amd.
Except for the one time it was not doing anything, other instances i could fix it with q-flashing bios (+backup) after i did regular update (first),
so nothing related to the cmos battery (but the chips/data).
 
Last edited:
Pretty nasty issue with these Gigabyte x570 boards. I have an X570 Aorus Elite and have not seen the problem. I've been on F11 since day one, been using the computer a couple months now.

It might be related to the dual BIOS, that's one of the big differences between the Elite and Master. Might be a good idea to disable automatic BIOS selection with the dip switches.

One thing I do generally is disable fast boot in Windows power options. I also disable hibernation with "powercfg -h off" at the command prompt. Might help for boards experiencing this issue.

From what I understand the issue commonly occurs after waking from sleep. It might be worth a try to set the sleep interval to zero (disabled) and see if the problem occurs.

There's some conjecture it may be related to the wifi card. It might be worth a try to disable it if not used. I believe the wifi card can also be removed. Otherwise if wifi is needed possibly go into advanced device settings from Windows device manager and disable any power saving or sleep modes if available.
 
Pretty nasty issue with these Gigabyte x570 boards. I have an X570 Aorus Elite and have not seen the problem. I've been on F11 since day one, been using the computer a couple months now.

It might be related to the dual BIOS, that's one of the big differences between the Elite and Master. Might be a good idea to disable automatic BIOS selection with the dip switches.

One thing I do generally is disable fast boot in Windows power options. I also disable hibernation with "powercfg -h off" at the command prompt. Might help for boards experiencing this issue.

From what I understand the issue commonly occurs after waking from sleep. It might be worth a try to set the sleep interval to zero (disabled) and see if the problem occurs.

There's some conjecture it may be related to the wifi card. It might be worth a try to disable it if not used. I believe the wifi card can also be removed. Otherwise if wifi is needed possibly go into advanced device settings from Windows device manager and disable any power saving or sleep modes if available.
Both times, mine has happened from when the computer was shut down properly and powered off normally the night before. So it's not related to just sleep or hibernate issues unfortunately.
 
I just joined these forums after being a somewhat lurker occasionally trying to find solution to my computer/tech issues after trying other possible solutions.
I noticed I was having issues like everyone else that has a Gigabyte X570 Aorus Master motherboard. Mine is paired with a Ryzen 7 3700x (stock) and 32GB GSkill Trident Z Royal 3600mhz (CL19).

I first had issues with my audio popping in games like FarCry 5, when swimming underwater in the game. I though I fixed it by changing the XMP High Freqency Support from Auto to Level 3, but it came back later.
Another issue that occurred like others in this thread is my BIOs kept resetting to default values, including my Windows OS time. I kept thinking I did not have this issue before since I purchase the X570 Aorus Master at product launch in July 2019.
I tried doing the remove CMOS battery trick for 10 minutes without power while pushing the clear CMOS button on the back of the motherboard twice, unfortunately it did not work for me.

I think the F11 BIOS has issues, at least with X570 Aorus Master and possibly others like the Elite & Ultra. I only upgraded to the F11 BIOS a couple of weeks ago seeing a new AGESA update, but my problems started showing after that.
I successfully downgraded both BIOS (main & backup) to F7b (released 09/17/2019) which adds update AGESA 1.0.0.3 ABBA .
I noticed that after doing this my XMP profile stays saved and my BIOS does not keep resetting on cold boots. The F11 BIOS has some regression bug that is corrupting the save bios settings is having to reset every so often.
I'm sticking with F7b BIOS for now as it was the first BIOs to support fan profiles for the chipset fan (PCH_Fan) in the SmartFan5 (F6) BIOS fan settings. Hope this helps others with this issue, also turn off PCI Express Link State Power Management and USB Selective Suspend settings in WIndows Advanced Power settings.

The Fix: Downgrade BIOS to F7b using in BIOS QFlash (released 09/17/2019)
 
Last edited:
CraigHB
neither fast boot nor hyber on on my rig, still same probs.
and only one or two boards have switches, others not. gets auto restored if there is an issue.

nanite84
i had problems on 6 already, cant run anything prior 10 if i want to use win, so not an option for me (but also shows F7 is not the fix, at least for ultra).
im not using xmp since 11, still, no change.


i did update to f10/11/12 incl backup bios, loaded defaults, "recovered" the same update in qflash mode (rig off), updated normal way again to overwrite data, loaded defaults,and it didnt fix anything.
since this is not just on amd, but intel too, so type of cpu/chipset etc can all be different, yet experience the same issues.
to me, that means its the hw or its implementation (of dual bios), and thus not something fixable "by us").
 
Last edited:
CraigHB
neither fast boot nor hyber on on my rig, still same probs.
and only one or two boards have switches, others not. gets auto restored if there is an issue.

nanite84
i had problems on 6 already, cant run anything prior 10 if i want to use win, so not an option for me (but also shows F7 is not the fix, at least for ultra).
im not using xmp since 11, still no change.


i did update to f10/11/12 incl backup bios, loaded defaults, "recovered" the same update in qflash mode (rig off), updated normal way again to overwrite data, loaded defaults,and it didnt fix anything.
since this is not just on amd, but intel too, so type of cpu/chipset etc can all be different, yet experience the same issues.
to me, that means its the hw or its implementation (of dual bios), and thus not something fixable *by us).
I agree it's not something we can fix on our end. I noticed for me I had less issues with the September 2019 bios. Gigabyte with American Megatrends has to fix these bios bugs, especially since we are paying a premium for these features.
 
Well, it just happened to me again. A little over a month since the last time. I went to turn it on and nothing. Pulled the battery, let it set for a few mins, cleared CMOS and reinstalled a brand new battery and she booted right up. I know the battery probably has nothing to do with it, but figured I might as well replace while I had it out. I will keep you all posted. This is the 3rd time since F11 BIOS was installed, nothing prior to that, which does lead me to believe it's something in this F11 BIOS that is soft bricking our machines.
 
Hi all,

New to the forum, and been drawn to this with the issues you guys are getting. My issue is slightly different but wasn't sure whether to make a completely separate post but here it goes.

I had a Asus Rampage V Edition 10 with a i7-6850k but it became unreliable as the power kept shutting off. I googled for weeks and tested many different theories but all came up short. I borrowed a PSU of a friend but same issue. Also started having an issue with USB port on a microphone. One day, i got so fed up with assuming it's another Asus shorting issue that after some research i brought the Gigabyte X570 AORUS Master with the AMD Ryzen 9 3900X combo and also using the stock CPU cooler that it came with. Set it all up but after 2 days i have had the same power issue when the PC trips but all the mobo led's etc stay on. It's only happened once at the time of writing but getting fed up. I didn't see what the error code was yet. Is it best to invest in a new PSU? I have updated the BIOS when i first built it to F11 but after the power issue it seems to have gone back to F5?

Spec:

AMD Ryzen 3900X with stock AMD RGB cooler.
Gigabyte AORUS MASTER X570
Asus STRIX GTX 1080 O8G
G.Skill Trident Z RGB 32GB (8GB x 4) 3200mhz PC4-25600 CL14
Gigabyte AORUS M.2 PCI-E 4.0
Samsung 960 Evo M.2
Samsung 860 Evo SSD
Samsung 850 Evo SSD
EVGA P2 850W Fully Modular

Any ideas please let me know.

Thank you in advance.
 
No, I didn't have any issues with previous BIOS versions. You may be on to something here.

I had this problem with previous bios revisions. More so in fact.

I've had this problem once in the last 6 months. I noticed that differing types of ram and timing caused me more problems. They were "stable" according to various stability programs but my booting got worse after tighter timings. I got rid of all my non samsung b die (I got rid of one of my pc's and sold my hynix sticks with it). I also detuned my timings. Since then I only had this problem once.
 
Found info this might be due to a beta bios "burning" its data into the chip,
causing all following updates to get messed up by leftover old stuff.
no difference if cmos was cleared or even if battery removed.

anyone willing to try:
flash bios down to the first 1x beta (has a letter after the digits) --> load defaults (change only what's needed to get back into bios after reboot),
flash bios incl backup the lowest (double digit) beta, reboot, load defaults and boot to win.
power down and qflash (recover) thru button and the special usb port on the back. after its finished, power up,
and flash the same bios AGAIN incl backup, and load defaults, shutdown.
if you want, remove cmos battery for 20 min, then turn on, load defaults, reboot, change to values you want.


Update:
rig reboot crashed again, so this seems to not make any difference.

BUT try to set pci speed to gen 3, as there seems to be a problem, especially when running dedicated (asus mainly) soundcards.
 
Last edited:
either im super lucky or just got a good one but i have not had a single issue with the x570 gigabyte master since installing it 08/2019. even the first iterations of the bioses i never had ram or cold boot issues. this is in stark contrast to the x470 gaming 7 wifi i came from. that board took 5 or 6 bios updates before it became compatible with my ram. currently running a 3800x and 4 sticks of 3600 ram with xmp.
 
either im super lucky or just got a good one but i have not had a single issue with the x570 gigabyte master since installing it 08/2019. even the first iterations of the bioses i never had ram or cold boot issues. this is in stark contrast to the x470 gaming 7 wifi i came from. that board took 5 or 6 bios updates before it became compatible with my ram. currently running a 3800x and 4 sticks of 3600 ram with xmp.
To be honest, aside from the 2 times mine has needed the CMOS battery pulled to resurrect it so far since January, it’s been a super solid board.

For me, it’s a minor pain in the ass at this point, but imagine being a system builder that put a system together for a customer with this issue- you’d have to RMA it for a different board especially for someone that isn’t tech savvy / willing to occasionally pull their vid card to do the procedure or refuses to accept that it’s “working properly” (rightfully so) as Gigabyte seemed to want to tell me when I put in a ticket about it.
 
but imagine being a system builder that put a system together for a customer with this issue

For an OEM that would be absolutely unacceptable, something worthy of a lawsuit for breech of contract. We consumers don't rate. It's like, "oh we have a fatal flaw with our product that we may or may not correct." What is that, the worst quality issue in the history of quality issues.
 
BUT try to set pci speed to gen 3, as there seems to be a problem, especially when running dedicated (asus mainly) soundcards.

Mine is already set to Gen 3 and has been since day one, so I don't think it's that either. At least we are crossing off the things it's not lol.
 
Im planning to sell cpu once the 4xxx series is out, but with the board action like this ill probably get rid of it too.
I know asus isn't always the best on service (not for me so far, even fixed a board i killed with a cheap psu and oc), but so far in 20y not one failed or even acted like my Ultra does.

Since the two x570 from MSI (cheapest ones btw) would even POST with the gskill ram (1st time powered on), while the Gb refused to do so,
unless i had set dram v and clock manually with some corsair sticks before swapping for the gskill kit (and tried 2 different ones).

Asrock always worked for me when the came on the market, but seem to go cheap on power design recently,
so i probably stick with a big MSI board (need 3 m2 slots) and an Asus as fallback.

I dont mind to miss a bit performance or have to do minor tweaks (as when Gb changed from samsung (rev 1) to hynix vram (1.1),
that needed more v that wasnt programmed and caused issues when the clocks were raised, but underclocking vram by 4-5 mhz would help),
but its one thing to change stuff because of changes with suppliers, and deliberately switching stuff to make more money,
without telling anyone or even noting it on product info etc.
Knowing this, i doubt Gb is really interested in doing anything.
Same reason no car rental company is paying for more than basic insurance, even if you pay for more.
the money they make outweighs the few times a car actually gets totalled.

Just done playing beta tester for companies, and just stay away from their products, and others will do,
and hopefully declining sales will have an impact.

Either way, i wont spend more time tying anything (unless i have to),
and since the reboots usually only happens when using office and/or chrome ( and even then only 1-2 times a month),
i can use it for the time being.

good luck everyone
rant mode: OFF
 
Hi everyone,

I have a similar problem but on the completely new mobo. Cheapest x570 from gigabyte: Gigabyte X570 UD.

At first, I tried to set up everything, powered the new build on, but nothing happened at all: no lights, no sounds, no fans.

Then I took the mobo out of the case, disconnected everything. Only ATX24 and 8-pin CPU cables left. Also CPU and CPU Fans.

Still did not turn on.

I found this topic on the forum (thanks a lot for sharing all the info!). Unfortunately, the trick with the CMOS battery did not work: I tried to leave it for 15 minutes, board completely unplugged. Did a CMOS reset a couple of times (on this board you need to short 2 pins to do it).

Unfortunately, it still did not help.

Shorting 4th and 5th pins on ATX 24 (when plugged to mobo) actually turns on all the fans: CPU fan, and mobo fan as well. But it does not turn on the PC.

Any ideas about what I could also try here?

Thanks a lot for any help!
 
Hello everyone,

Like several others in this thread, I made an account specifically for this thread.

I'll try to keep this concise as most of what I've done or experienced mirrors the majority of people replying to this.

I built this PC in November 2019 and carried over only my graphics card and power supply from my previous PC. Everything else was brand new hardware including, of course, the Gigabyte AORUS X570 Master. (Came with BIOS F10. Still using F10. Flashed to F11 at one point but after experiencing this issue my board reverted to F10 and I have not bothered to reflash it since).

On February 5th, 2020, I first experienced this no-start up issue. I would press the power button and nothing happens except for all lights on the board shutting off (namely the built-in power button). Unplugging the PSU/powering it off, and trying the CMOS reset button on the back of the board doesn't do anything. No fans, no POST. I shut down my PC after I'm done with it daily. I never use sleep mode, and I've long since disabled the applicable Windows 10 settings that have been suggested to be a probable cause. At first I suspected the Power Supply, especially since I had carried it over from my old PC, so I ordered a replacement. Two days later I go to install the brand new PSU and to my extreme frustration the issue is persisting. After various steps of troubleshooting (Mind you this is AFTER I'd written off CMOS as a potential cause because I tried the CMOS reset button on the board at least twice) I just decided to do a sanity check and eject the cell. I waited 15 seconds with the power button held down, CMOS cell ejected, and power disconnected, put the CMOS cell back in, and the machine powered right up. At that point I was just relieved and went on with my day.

On February 29th, 2020, the same exact issue occurred. I ordered a pack of Energizer CR2032s and they arrived on March 1st. In the meantime I once again ejected the cell (which, for any potential buyers, requires removal of the graphics card to access if the card is pretty large) and reinstalled it to last me until it potentially happened again.

As of now, May 25th, 2020, the issue seems to occur about once a month for me. The third occurrence I replaced the CMOS cell that came with the board with one of those Energizers, and as of last night, May 24th, 2020, I now have the third CMOS battery on the board since I got it in November last year and I anticipate this issue will never go away for as long as I have this board. I'll be contacting Gigabyte sometime today and will reply if I hear anything back other than the usual reply everyone else seems to get.

ALSO, in case this may help narrow down a cause, this most recent occurrence of the issue I shut down my PC, I noticed the LEDs in my Corsair K95 RGB Platinum stayed on in their default lighting configuration after shutdown, but only the keyboard. Currently it's plugged into two of the rear USB ports on the motherboard with it's built in passthrough not being utilized. When I saw this I attempted to power the PC on so I could shut it down again so that my apartment wouldn't be glowing red while I was trying to sleep and that's when the X570 AORUS Master issue resurfaced it's ugly head again.

System specs:
Gigabyte X570 AORUS Master
AMD Ryzen 9 3900X (Default clock)
EVGA GeForce RTX 2080 FTW3 (Using EVGA factory OC settings)
Corsair CMW32GX4M2C3200C16 VENGEANCE RGB PRO 64GB (2 kits of 2x16GB) DDR4 3200 (PC4-25600), running XMP @ 3133MHz since I wasn't thinking straight when I bought two separate kits of non QVL memory at their advertised speed.
 
Last edited:
Hello everyone,

Like several others in this thread, I made an account specifically for this thread.

I'll try to keep this concise as most of what I've done or experienced mirrors the majority of people replying to this.

I built this PC in November 2019 and carried over only my graphics card and power supply from my previous PC. Everything else was brand new hardware including, of course, the Gigabyte AORUS X570 Master. (Came with BIOS F10. Still using F10. Flashed to F11 at one point but after experiencing this issue my board reverted to F10 and I have not bothered to reflash it since).

On February 5th, 2020, I first experienced this no-start up issue. I would press the power button and nothing happens except for all lights on the board shutting off (namely the built-in power button). Unplugging the PSU/powering it off, and trying the CMOS reset button on the back of the board doesn't do anything. No fans, no POST. I shut down my PC after I'm done with it daily. I never use sleep mode, and I've long since disabled the applicable Windows 10 settings that have been suggested to be a probable cause. At first I suspected the Power Supply, especially since I had carried it over from my old PC, so I ordered a replacement. Two days later I go to install the brand new PSU and to my extreme frustration the issue is persisting. After various steps of troubleshooting (Mind you this is AFTER I'd written off CMOS as a potential cause because I tried the CMOS reset button on the board at least twice) I just decided to do a sanity check and eject the cell. I waited 15 seconds with the power button held down, CMOS cell ejected, and power disconnected, put the CMOS cell back in, and the machine powered right up. At that point I was just relieved and went on with my day.

On February 29th, 2020, the same exact issue occurred. I ordered a pack of Energizer CR2032s and they arrived on March 1st. In the meantime I once again ejected the cell (which, for any potential buyers, requires removal of the graphics card to access if the card is pretty large) and reinstalled it to last me until it potentially happened again.

As of now, May 25th, 2020, the issue seems to occur about once a month for me. The third occurrence I replaced the CMOS cell that came with the board with one of those Energizers, and as of last night, May 24th, 2020, I now have the third CMOS battery on the board since I got it in November last year and I anticipate this issue will never go away for as long as I have this board. I'll be contacting Gigabyte sometime today and will reply if I hear anything back other than the usual reply everyone else seems to get.

ALSO, in case this may help narrow down a cause, this most recent occurrence of the issue I shut down my PC, I noticed the LEDs in my Corsair K95 RGB Platinum stayed on in their default lighting configuration after shutdown, but only the keyboard. Currently it's plugged into two of the rear USB ports on the motherboard with it's built in passthrough not being utilized. When I saw this I attempted to power the PC on so I could shut it down again so that my apartment wouldn't be glowing red while I was trying to sleep and that's when the X570 AORUS Master issue resurfaced it's ugly head again.

System specs:
Gigabyte X570 AORUS Master
AMD Ryzen 3900X (Default clock)
EVGA GeForce RTX 2080 FTW3 (Using EVGA factory OC settings)
Corsair CMW32GX4M2C3200C16 VENGEANCE RGB PRO 64GB (2 kits of 2x16GB) DDR4 3200 (PC4-25600), running XMP @ 3133MHz since I wasn't thinking straight when I bought two separate kits of non QVL memory at their advertised speed.
Thanks for letting us all know. It seems to be the same issue that I have had. Someone in another forum said to Enable ERP in the BIOS, so I have done that in hopes it might be a solution. Please keep us posted.
 
here a list with all updates for all boards, helps if you can read german, but not needed.
aega overview
when u look at Gb, it says they are working on an update.
funny thats its only on the boards that (according to Gb) dont have problems :rolleyes:

USER915130
unplug power cord and remove cmos batt, press AND HOLD power button on case for 10s,
then leave it overnight.

FlyingToastr
the kb having lights on after shutdown is a common thing ive seen on all boards that offer usb power
when system is off (to charge phone etc).
all my msi/asrock/asus boards in the past had a bios setting, with Gb its done thru the stupid win app,
so usually nothing you can do.
since i (usually) dont sleep at my desk, i just unplug/plug the kb back in, after i get up from the chair, if the kb stays lit.

update: seems to be caused by win, as the latest build (2004) doesnt leave the lights on, nor do they turn on if i hit a key (while pc is off).
 
Last edited:
yeah, same here. System not posting once every week after a shutdown. I have the x570 aorus master and 3900x. When that happens, the little red light on the mobo turns on when the board has power, but once I press the power on button it goes away and nothing happens, as if the system is dead. Removing and then putting back again CMOS solves the issue temporarily.
Have you guys had any luck contacting Gigabyte? Do we know if this is some sort of hardware issue that needs to be RMAed or something that can potentially be fixed with a BIOS update (I'm currently running latest, F11)?
 
yeah, same here. System not posting once every week after a shutdown. I have the x570 aorus master and 3900x. When that happens, the little red light on the mobo turns on when the board has power, but once I press the power on button it goes away and nothing happens, as if the system is dead. Removing and then putting back again CMOS solves the issue temporarily.
Have you guys had any luck contacting Gigabyte? Do we know if this is some sort of hardware issue that needs to be RMAed or something that can potentially be fixed with a BIOS update (I'm currently running latest, F11)?
Was basically told by Gigabyte support that if this was a widespread issue that they would have thousands of reports, and to go pound sand.

I’ve been “lucky” so far and it’s only happened twice to my rig.
 
I encountered another issue--such fun...;)

1) As I mentioned in previous posts, the cold-boot failure (reporting no boot drive available) I was having with my 3900X & Aorus x570 has not returned--not even once--since I replaced the (evidently faulty) pwr strip I had been using. So that's good.

2) Also, I have not seen the faux dead motherboard situation again since the first time. That's good, imo...;)

2)Instead, what was happening now is that every cold boot the bios was resetting--and a couple of times I noticed the clock losing time, again. Not so good. Unlike before, however, the system would boot to desktop with ram @ 1066MHz instead of 1867MHz. It indicated a problem with the ram and the XMP 2.0 overclocking on the mboard.

3)I decided to try the F12f beta bios, which I am still running. No dice--every cold boot the system boots to desktop, but the ram settings are lost and have to be reloaded/reset.

I was kind of desperate by this time--but I was determined I was not going to pull and/or replace the battery again--something was causing this problem and obviously the drained battery was a symptom rather than the cause.

But what the heck could the cause be, then?

Out of sheer desperation I took the suggestion of USAF46270 above about enabling ERP again in the bios and I enabled it even though turning it on and off beforehand had made no difference in my original problems.

So...ERP ON in my F12f bios install seems to have made the difference. From the first cold boot with ERP On the system XMP settings are rock solid. No battery drains--didn't have to pull it, etc. Three successive cold boots without a ram bios reset are behind me! I hope it holds! Of course I'll be back if it doesn't.

[All of this is counterintuitive to me! Doesn't make much sense to me..;) How could ERP ON serve to improve ram training? Beats me. But everything is working fine now with ERP ON and no battery pull proved necessary, thankfully.]
 
[snip]
[All of this is counterintuitive to me! Doesn't make much sense to me..;) How could ERP ON serve to improve ram training? Beats me. But everything is working fine now with ERP ON and no battery pull proved necessary, thankfully.]
I don't think it's a case of "improving ram training", but rather now your system isn't losing the existing ram training it had already previously learned (and saved) from the first boot with that RAM installed (plus any XMP settings/changes that you applied in the BIOS after that first boot).
As you know, the BIOS saves the timings and sub-timings for the installed RAM as config settings the (very) first time it boots up or if the BIOS thinks the memory or CPU changed since last boot.
Whatever this motherboard's problem is, we know that messing around with (removing) the CMOS battery fixes it pretty reliably.
This tells me that the non-boot problem which this thread is all about AND your specific problem with clock being wrong and RAM speeds resetting are all somehow related to how and what the motherboard BIOS or other low-level firmware is doing to the memory that is backed up by the CMOS battery.
Something is screwing around with the data stored in that memory and corrupting/clearing it "randomly" on shutdown (I think).
By enabling ERP, you've changed the behavior of the very low level firmware (lower level than BIOS mind you, since it controls stuff like Wake-On-Lan, Wake-On-Keyboard, etc.) so that it is no longer attempting to power certain components (LAN controller) and USB devices (keyboard) while the system is off. From your experience so far, enabling ERP might help reduce or eliminate the cold-boot "won't turn on" problem and your "BIOS forgot settings" problem.
So yeah, I'm beginning to think that your BIOS forgetfulness issue and the cold-boot won't power on issues are both related to the same underlying design or low-level component/firmware problem that this the subject of this thread.
 
ErP is a new one for me. My last system did not have it. Had to look up what it does and it's something I want. I want the computer to be as fully off as possible in s5 shutdown and ErP enabled does that. So I enabled it.

In general the Gigabyte x570 BIOS seems to have some odd bugs. Changing settings can cause seemingly unrelated problems. For example on my x570 Elite, if I disable CSM the BIOS becomes laggy when navigating and I get a black screen for a longer period of time on boot. I want to run with CSM disabled, but it has side effects to I have to leave it on.

I think that's inexcusable when basic switches in the BIOS cause problems like that. I have other complaints with the BIOS and I'm super unhappy with it, enough that I will not go with Gigabyte again. I think they have a back room full of monkeys beating a motherboard writing their BIOS software.
 
  • Like
Reactions: Azrak
like this
ErP is a new one for me.
Me too, until this thread.
I found a decent article about what it is, why it exists, and what it means to have it enabled or disabled: https://www.onecomputerguy.com/erp-ready

The key thing is: When ErP is ENABLED, the computer will be really shut down as much as it can be. When I first heard about it, I figured having it enabled meant that the computer would be partially powered on for Wake-on-Lan, etc, but it is the opposite of that. Enabling ErP turns on the power SAVING features, so the computer uses less power when turned off compared with ErP disabled.

from the article said:
When you have ErP mode enabled and the computer is shut down, there is no saved state for the computer to return to, and a full boot of the system is required in order to restore working functionality.

ErP mode disables the following four settings when it is enabled:
  • PME Event Wake Up – PME is short for “power management event” and is responsible for devices—which do not have ErP enabled—turning on again spontaneously after being shut down by the user.
  • Power On By Mouse – You will not be able to return the computer to a wake state by using the mouse.
  • Power On By Keyboard – You will not be able to return the computer to a wake state by using the keyboard.
  • Wake On LAN – Incoming signals from the LAN will not cause the machine to return to a wakeful, working state.
The trace power consumed by the system does not power USB ports or other peripherals, and these devices will not be able to charge while the computer is in a shutdown state.
 
joined just for this thread. so i bought an aorus x570 pro wifi last week (refurbished) and had the issue where the computer would not turn on after i turned it off. i found this thread and disconnected the cmos battery, reconnected the battery and put the computer back together (i have to remove the mobo to get to the battery). pressed the power button and it still didn't turn on, so back to the internet i went on my phone looking for a solution. a couple of minutes later i hit the power button again and it turned on. so for the past week i've been using this setup but i've had to press the power button a couple of times before the computer turns on, i can live with it.

then today i decided to switch from my atx case to my matx case and i could not get the computer to turn on in the matx case. i transferred everything over, same cables and components (not the case fans), disconnected cmos battery, cleared cmos. so after a while i finally gave up on the matx case and decided i would just return the motherboard. right before i boxed the motherboard up, i decided to put everything back in the atx case just for the heck of it........and it runs. still takes a couple of presses of the power button but it runs. so i'm not sure if i want to keep this mobo or return it before the return period ends but its working now.

i'm using a 3800x and had the same issue on f3 bios before i updated to f10.
 
i guess i have to state this again:
there is no "DISCHARGING" issue with those batteries, as they are NOT rechargeable, ergo if it REALLY was a discharge, putting the empty battery back in,
would make the board would at least behave as before.
which also means, unless you actually measure the voltage and the battery is really empty (age etc), replacing the battery multiple times will NOT fix anything.
the fact that the bios loses only the ram settings but keeps other things, again says, this is NOT an empty battery issue (there is no "a little bit pregnant" :D )
not saying that ppl cant have an empty battery, but definitely not causing +90% of the issues on Gb boards posted here.

ErP is only a temporary thing, and for ryzen builds NOT recommended at all, because of the way the cpus work.
The main reason is to have it turned on, to limit power consumption in soft-off state and be able to power up rig with kb/mouse/Ethernet event.
erp bug

Do say try the following:
make sure to run the latest bios, no matter if bios or not, set everything needed for 1usmus power plan except the psu to low load (instead of typical load),
dont use PBO (use auto or OFF/disabled) dont use xmp, just set clcok/timings/voltage manually, turn off gear down/power down for ram,
disable spread spectrum, enable HPET, switch CSM to enabled (should be on for most GB boards anyway).
turn any memory training to OFF (worst case clear cmos), shouldnt be an issue if u set clocks/timings that worked before.

I had a reproducible cold boot after rig was cut from power (either psu switch or pulling plug), that now does not show anymore,
so far no sudden reboots or anything like memory retraining happening.
under 1909 i had that, so far not happening under 2004 build.
if you want to try it, i recommend a clean install (make sure bios is set to UEFI not "other OS" prior to install),
and use the latest drivers incl the amd (not 1usmus) balanced plan.


Win 10 2004 release
Its seems MS did a couple of tweaks/optimizations towards ryzen, as i now see the cpu clock dropping below 3.6 ghz (2.85) on amd balanced plan,
which it never did before, clocks seem to be less jumpy, and even that i know the latest bios (beta) has made no difference to cold boot or sudden reboots,
since running the 2004 build i had no issues.
(and this with the same cmos battery that the board came with)

That said, do NOT install/update (win 10) to 2004, as they found issues with wireless and other things,
and are blocking affected computers from the update server on purpose, until after the fix.

Do try IF you can make a proper image of your OS drive (with Ghost or True Image), as it seems to be working so far,
at least when combined with latest bios (even if beta) and drivers, while using amd power plan (1usmus has not update his since nov 2019, but bios/drivers got updated twice, so i will hold back on using it until its up-to-date anyway).



1366_warrior

if you can, yeah return/swap for either MSI, or maybe an Asus. Asrock boards are good hw wise, but i remember some issues/missing settings in bios,
that made me drop it (for now) as an option for amd boards.
 
Last edited:
I've been up and running with this build since mid December and this has happened to me only once with my Gigabyte Aorus pro wifi in March or April. MB was completely dead until i pulled the battery. I believe i was on bios ver F10. Not a big deal so far i hope it doesn't become an issue.
 
Just wanted to mention: this was an issue on some of their (Gigabyte's) X370 boards as well. My board (a gaming k5) didn't have it, but quite a few members of the gigabyte forum had reported it. I'm not aware of whether it ever got fixed.
 
and this is the response I got back from Gigabyte:

Dear sir,

We recommend you to apply for warranty at the local vendor you've bought the motherboard from, since it's the only way it can be properly diagnosed.

Best regards,

GIGABYTE team

Go figure that out... they want me to send it out for service just to get it back 2 months later with the same or worst condition.
 
Guys, does replacing the CMOS battery with a new one and turning ErP on (or even updating to F12x BIOS) help make reoccurrence of the issue less frequent? If it's like once per month or so I can deal with it, even if it means replacing battery and everything. However, I'm not willing to send the mobo for service as I need my system for my work.
 
Nope.
Im running the same battery it came with, put in a new one (temporally) and even an older one (with less voltage), no change in behavior.
Do have less issues since the F12e beta, only sudden reboots, no cold boot issues anymore.
 
ErP is only a temporary thing, and for ryzen builds NOT recommended at all, because of the way the cpus work.
Do you have a source for this information?

Nope.
Im running the same battery it came with, put in a new one (temporally) and even an older one (with less voltage), no change in behavior.
Do have less issues since the F12e beta, only sudden reboots, no cold boot issues anymore.
Why are you saying "Nope"? You said don't have the cold boot issue. Is ErP enabled on your system?
 

Follow-up to my original reply. Gigabyte replied in a rather unsurprisingly unhelpful manner:

Dear customer,

Seems you have tested everything. Please check one more thing, remove cpu from MB, inspect if any burned mark / bad pin on cpu / socket found. If cpu socket look fine, re-seat cpu back, make sure mount cpu cooling on top tight, connect cpu cooling to cpu fan header, test run system with single stick memory a time on slot 1 ( slot close to 24 pin power connector ) to rule out if one of the unstable memory cause issue. Try test with other graphics card as well.

If issue still persists, the next step you will need is whether refer to Gigabyte RMA service let our MB specialist examine the complete board for you.

Gigabyte MB carries an 3 year warranty, log on http://rma.gigabyte.us. Go to user support, register account to Gigabyte, complete and submit an RMA request form. RMA number and shipping instructions will be assigned 24-48 business hours via e-mail for you to send the motherboard to RMA dept for service.

Best regards,

GIGABYTE technical support team.

My reply to them was

There are no burn marks or bad pins on the CPU or socket. I will not waste time trying to test one stick of memory. As I stated in the original service request the system runs completely fine for around a month (with multiple shutdowns and restarts throughout) before the issue reoccurs. What you're asking me to do is run the computer for four months with bare minimum memory, rotating sticks out once every 30 days, and wasting four months of my time to chase an unlikely cause for the issue. Additionally, while I'm flattered that you feel like I can afford a 2nd video card (and to run that 2nd video card for a month in hopes that the issue does not reoccur), I regret to inform you that that is simply not the case. I also cannot afford to go without my computer for extended periods of time, nor can I afford a spare motherboard while your service centers keep my motherboard for 30 days to try to replicate the issue.

I sincerely doubt anything useful will come of this. ASUS is looking very tempting right about now. I'll reply again if I get anything useful out of Gigabyte, but I wouldn't hold my breath.

EDIT:
Gigabyte replied to the above and basically just said I need to RMA the board and to expect a 7-10 day turnaround. I'm going to call them and ask for any potential verification that there is a hardware revision that might resolve the issue. My best guess is that if it is not a BIOS issue, there may be crosstalk between some hardware on the board that's causing bad data to be written to CMOS upon shutdown. Without a hardware revision I have no confidence that a replacement will solve the issue, and even less so that their RMA technicians will be able to replicate the issue within 7-10 working days even with their oscilloscopes.

If they state there has not been a revision of the board since I got mine, or won't say, I'm just going to deal with this issue until X670 (assuming that's what it will be named) gets released and then get an ASUS X670 board. Gotta say, after having used Intel since Pentium 4 HT, its nice to have a CPU that can be used on the next generation of boards (once again an assumption.)
 
Last edited:
Was basically told by Gigabyte support that if this was a widespread issue that they would have thousands of reports, and to go pound sand.

I’ve been “lucky” so far and it’s only happened twice to my rig.
And I thought that ASUS support was terrible.
 
Azrak
i say nope because i have done all kinds of "tests" with different batteries in different states
that i can say it has ZERO effect on the problems.
its like asking a (proper) driver, that had a car break down, if they ran out of gas. sure, there might be one or two that forgot, but +90% of the time it will not be the cause.
if i have a board showing all those problems, no matter if the cmos battery is almost empty/half/fully charged, i can exclude it as the cause.
and any that have a "discharge" problem, there is probably a defect on the board, or the battery (unless very old) would not get drained in the first place,
certainly not on boards less than 3y old.


ErP is a EU requirement to SAVE power when in sleep/soft off state
and its actually the complete opposite of what we need for cold boot issues. your increasing the difference between power off to power on,
it will probably cause more, rather than less cold boot problems, as it acts like flipping the psu switch (which for me causes a cold boot EVERY TIME)
not saying NO one will see this help, but again, for +90% its not helpful to turn it on.
i dont remember the site regarding ryzen (as i found more than one and for me its important to retain the info, not the source; using google doesnt make you lose your eye sight.. :D),
but how about ppl actually have problems AFTER turning it on?
for me, it completely disabled wifi/bt, and only after disabling ErP AND leaving rig unplugged for a bit made it work again.
erp bug
GB uk

i will never say dont try something if your rig has issues, but i have a problem with ppl recommending things
1. without knowing what is for and/or what it does (and the like)
2. never verify with "A/B" testing or by elimination (etc) if their theory is correct.

as example one guy claiming you shouldn't touch the psu idle current setting and leave it on auto.
fact is, it only has 2 states ("typical" or "low"), which means even on auto, the bios will have to select one or the other,
and he could not explain or even name just a reason (without links/reference) why (auto).


now running 2004 build for a while, and when using "tweaked" (towards ryzen) bios settings, i have not had any issues,
the only thing left would be the sudden reboot crashes, but they sometimes dont show for 2-3 weeks, so might still show later.

Update June 10:
had 2 sudden reboots about a week apart, tho nothing else acting up.
 
Last edited:
Back
Top